1. Universal Determinism and Deterministic Sys-
tems. Determinism, so say determinists, is misun-
derstood and misrepresented by its adversaries; and
nowhere more than in historiography. Although anti-
determinists justly retort that their position has fared
no better, they cannot well deny the determinists'
complaint. The English word “determinism,” like its
French, German, and Italian counterparts, is of seven-
teenth- and eighteenth-century coinage. It was intro-
duced as a name for two different, but related, doc-
trines. One, the doctrine that choice between different
courses of action can, in all cases, be fully accounted
for by psychological and other conditions, has as yet
played little part in historiography. The other, which,
to avoid ambiguity, may also be called “universal
determinism,” is the doctrine that everything that
happens constitutes a chain of causation, a doctrine
which obviously implies that human history forms part
of such a chain.
Universal determinism depends on a concept of
causation that was not generally adopted until after
the seventeenth-century “scientific revolution.” In an-
cient and medieval philosophy, a cause was conceived
simply as that which produces an effect. Some causes
were taken to produce their effects necessarily, as a
moving hand holding a stick necessarily moves that
stick. Others were taken to have the power to produce
an effect, which they might exercise or not without
necessitation, as a man without necessitation exercises
or does not exercise his power to move his hand. Most
ancient and medieval philosophers accepted the prin-
ciple that every event has a cause. But since most of
them took some happenings or events (namely, human
or divine actions), to be caused by agents and not by
other events, they held that some causes (namely,
human or divine actions), are not themselves events
in a causal chain. Hence they were not determinists.
The Greek atomists suggested another feature of the
concept of causation, which the work of Galileo,
Descartes (even though he was not an atomist), and
Newton was to establish in natural science. On this
concept, every event in nature is a stage in a process
the course of which is determined by laws of nature,
and can be considered a necessary consequence, ac-
cording to those laws, of earlier stages in that process.
A cause of an event is simply a set of initial conditions
that are, according to laws of nature, jointly sufficient
for its occurrence.
This concept of causation underlies Laplace's strik-
ing formulation of universal determinism, in the Intro-
duction to his Essai philosophique sur les probabilités
(1814). Treating the history of the universe as a single
process, he maintained that, from a complete specifi-
cation of the state of the universe at a given instant
(initial positions and velocities of all bodies), a super-
human intelligence knowing the laws of nature could
infer all past and all future states of the universe.
Laplace assumed that mass, position, velocity (the
terms of Newtonian physics) would suffice for the
required specification. Since it is doubtful, however,
not only whether the terms of Newtonian physics or
any possible future physics would suffice, but also
whether even a superhuman mind could specify, in any
terms, a state of the whole universe, Laplace's formu-
lation has been rejected by many determinists. It is
now more promising to define universal determinism
as the doctrine that every event in principle falls within
some deterministic system.
A deterministic system, in the sense here considered,
is a system of things in the universe. For any such
system there is a set of characteristics, each of which
is truly or falsely predicable of each thing in the system,
and some of which allow of variation in magnitude
or intensity (the variables of the system) such that a
state of the system is specified by a description of
everything in it in terms of all the characteristics in
that set. An event in the system may be defined as
any persistence or change in any of its states, in any
respect during a temporal interval. Such a determin-
istic system must, in addition, satisfy three conditions:
(1) all events in it must in principle be explicable
according to fundamental laws, which (2) mention no
characteristics except those in terms of which states
of the system are specified, (3) the explanations being
such as refer to no thing or event outside the system.
Bergmann has usefully labelled the second of these
conditions as “completeness” and the third as “clo-
sure.” Deterministic systems, in this sense, are inevita-
bly abstract. The solar gravitational system, for exam-
ple, consists of the sun, the planets, and so forth,
considered solely with respect to the characteristics
taken account of in gravitation theory, and not as
concrete objects. The duration of such systems is nor-
mally limited: thus, according to astronomers, the solar
system had a beginning, and will have an end. It is
a fallacy to infer that, because such systems are abstract
and impermanent, they are not real.
The complexity of a given deterministic system sets
a limit on how adequate a theory can be developed
of it. The Newtonian theory of the solar gravitational
system, which inspired Laplace's formulation of uni-
versal determinism, is almost uniquely adequate be-
cause the solar system is, in two respects, almost
uniquely simple: both the number of bodies composing
it—sun, planets, comets and so forth, and the number
of variables by which its states are defined, are com-
paratively few. Hence, it is practicable not only to
establish its state at the present time, but also, by the
Newtonian laws, to compute with reasonable accuracy
its past and future states. By contrast, it would be
utterly impracticable to attempt a similarly adequate
theory of the earth's geological history; for the geolog-
ical state of the entire earth at a given time would
be far too complex to define, and the variables deter-
mining geological change are more numerous. Geolo-
gists accordingly simplify. They explain geological
changes by constructing simplified models representing
states of the earth or of parts of it at different times,
and showing how, according to established laws of
nature, the forces at work within one simplified model
would bring about a transition to another. For more
complex systems, we must be content with even less
adequate sketches of a theory.
The concept of a deterministic system has led to
extensions in the meaning of “determinism” and its
cognates in the following way. One who maintains that
a system S is deterministic, or that a set of events K
falls within some deterministic system, is naturally said
to have embraced determinism with respect to S or
K. Such extended special usages are more common in
historiography than the general philosophical ones
hitherto considered. Thus Pieter Geyl has described
determinism as “represent[ing] the historical process
as a concatenation of events, one following upon the
other inevitably, caused as they all are by a super-
human force or by impersonal forces working in society
independently from the wishes or efforts of individuals”
(Debates with Historians, p. 238). He appears to have
in mind the view, accepted by not a few historians,
that social systems are, or are parts of, deterministic
systems, even if individual human actions are undeter-
mined.
Such views as that a given system of things in the
universe is deterministic, or that a given set of events
falls within a deterministic system, will hereafter be
referred to as “special determinist doctrines,” by con-
trast with universal determinism or the doctrine that
every event in the universe falls within some deter-
ministic system. Two elementary facts about the logical
relations between universal determinism and special
determinist doctrines are often neglected. First, uni-
versal determinism does not entail any special deter-
minist doctrine. In particular, universal determinism
does not entail the special doctrine which Geyl calls
“determinism”: it implies that human actions have a
place in the causal series, but has nothing to say about
what that place is. It is compatible both with the
doctrine that the wishes and efforts of individuals can-
not affect large-scale historical processes, and with the
doctrine that they can and do. Secondly, special deter-
minist doctrines do not necessarily imply or presuppose
universal determinism. Thus, the special form of deter-
minism mentioned by Geyl appears to allow that the
wishes and efforts of individuals may not fall within
a deterministic system. The classical example of this
logical independence, however, is found in the philos-
ophy of Descartes, who considered the system of mo-
tion and rest in the realm of matter (res extensa) to
be deterministic, except when changes in it were
caused by the activity of thought (res cogitans) which
he took to be physically undetermined. In a Cartesian
universe, even though virtually all happenings in the
material world fall within a deterministic system, uni-
versal determinism fails to hold for acts of the mind.
